Forecast Analysis of Instant Noodle Demand using Support Vector Regression (SVR)

Support Vector Regression (SVR) is a part of Data Mining (DM) techniques where it can be used for forecasting the instant noodle. The cycle of the product demand is hard to predict. It will influence the resistant of the product quality where the product be expired easily and the other thing is the market demand. The objective of this research is approaching the predictive models with their performance measured with Mean Square Error (MSE) of SVR The data was collected from the determinant of instant noodle demand dataset. The random normal generated data was explored to get the amount of specific data. Then, it used SVR to forecast the demand. The result of this study the MSE of standard is 1.612 and the SVR is 1.436, means it increases around 11% better the performance than the original dataset. Since, we conclude that the SVR method would be promising to be one of a forecast demand method.